GatewaystoHawai‘i:GeneticPopulationStructureoftheTropical ...downloads.hindawi.com/journals/jmb/2011/783030.pdf · Kingman reef are important centers of the genetic diversity in

Hindawi Publishing CorporationJournal of Marine BiologyVolume 2011, Article ID 783030, 16 pagesdoi:10.1155/2011/783030

Research Article

Gateways to Hawai‘i: Genetic Population Structure of the TropicalSea Cucumber Holothuria atra

Derek J. Skillings,1, 2 Christopher E. Bird,1 and Robert J. Toonen1

1 Hawai‘i Institute of Marine Biology, School of Ocean & Earth Science & Technology, University of Hawai‘i at Manoa,Kane‘ohe, HI 96744, USA

2 Department of Zoology, University of Hawai‘i at Manoa, Honolulu, HI 96822, USA

Correspondence should be addressed to Derek J. Skillings, [email protected]

Received 9 July 2010; Revised 2 October 2010; Accepted 3 November 2010

Academic Editor: Benjamin S. Halpern

Copyright © 2011 Derek J. Skillings et al. This is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properlycited.

Holothuria atra is one of the most common and widest ranging tropical, coral reef sea cucumbers in the world, and here we examinepopulation genetic structure based on mitochondrial COI to aid in determining the appropriate scale for coral reef management.Based on SAMOVA, AMOVA and BARRIER analyses, we show that despite its large range, H. atra has hierarchical, fine-scalepopulation structure driven primarily by between-archipelago barriers, but with significant differences between sites within anarchipelago as well. Migrate analyses along with haplotype networks and patterns of haplotype diversity suggest that Hawai‘i andKingman reef are important centers of the genetic diversity in the region rather than an evolutionary dead-end for migrants fromthe Indo-Pacific. Finally we show that for H. atra Kingman Reef is the most likely stepping stone between Hawai‘i and the restof the Pacific, not Japan or Johnston Atoll as previously presumed. Based on our data, Johnston Atoll can instead be seen as anoutpost of the Northwestern Hawaiian Islands rather than a gateway to the Hawaiian Archipelago.

1. Introduction

Echinoderms play a major role in structuring many marineecosystems, and many are described as “keystone species”because of their profound influence on benthic communitystructure (e.g., [1–3], reviewed by Uthicke et al. [4]). Inaddition to their important ecosystem functions, manyechinoderm species are also the focus of artisanal or com-mercial fishing efforts, particularly the sea urchins and seacucumbers [5–7]. The influence of echinoderm harvest ona wide range of other commercial fisheries, such as abalone,lobster, kelp, and kelp-associated fin fish, has long stimulateddiscussions of multispecies approaches to managing theirexploitation (e.g., [5], reviewed by Purcell [7]). Delineationof the appropriate spatial scales for management zoneswithin a spatial management network requires a detailedunderstanding of dispersal pathways and population connec-tivity (reviewed by Hedgecock et al. [8], Thorrold et al. [9],Fogarty and Botsford [10]).

Understanding connectivity in the sea is complicated bythe fact that many marine organisms share a biphasic life

cycle typified by an adult form that is relatively sedentaryand a larval form that can potentially disperse across largeexpanses of open ocean [11–15]. For example, in the seaurchin genus Tripneustes, some well-known biogeographicbarriers, such the Isthmus of Panama or the long stretch ofdeep water in the western Atlantic, are important barriers todispersal whereas others, such as the Eastern Pacific Barrier,show no evidence for limiting dispersal [16]. However, thegeographic limits of such dispersal are uncertain because itis virtually impossible with current technology to directlytrack these microscopic juveniles during the pelagic phase(reviewed by Levin [17]) making indirect methods ofquantifying larval dispersal particularly attractive (reviewedby Hedgecock et al. [8], Grosberg and Cunningham [18],Selkoe et al. [19], and Hellberg [20]). Proxies for dispersal,such as pelagic larval duration (PLD) and geographicrange, have generally been used as rules of thumb in theabsence of a detailed understanding of connectivity formost marine species. Unfortunately, intuitive expectationsof larval dispersal potential as a function of PLD and

2 Journal of Marine Biology

range size are not upheld in recent meta-analyses of theexisting literature [21–26, 88]. Realized dispersal distanceis typically less than potential dispersal distance because ofthe presence of biophysical or biogeographical barriers [27–31]. Barriers that limit dispersal between marine populationsinclude obvious geographical features such as land masseslike the Isthmus of Panama [32], but also more subtlefactors such as currents and oceanographic regimes [33–37]. The correlation between geographic distance and theprobability of larval exchange among sites is low in manymarine systems (e.g., [38]), and thus quantitative estimatesof connectivity are an important prerequisite for delineatingthe appropriate scale over which marine populations oughtto be managed.

The Hawaiian Archipelago lies at the periphery of thetropical Central Pacific and is the most isolated island chainin the world, making it biogeographically partitioned fromthe rest of the Pacific Islands (reviewed by Ziegler [39]).This isolation results in one of the highest proportions ofendemism in the world (e.g., [40–42]; reviewed by Ziegler[39], Eldredge and Evenhuis [43]). Though there are manyexamples of pan-pacific coral reef organisms in Hawai‘i, theisolation of the Hawaiian Archipelago is thought to limitlarval exchange sufficiently that colonization is rare [44].For example, Kay [45] estimated that Western Pacific marinespecies successfully colonize the Hawaiian Archipelago aboutonce every 13,000 years. Unlike the terrestrial fauna, how-ever, the Hawaiian marine fauna contains a large proportionof endemics that are differentiated but not diversified fromtheir Indo-West Pacific roots [39, 46–48]. Johnston Atoll isbelieved to be a stepping stone into Hawai‘i, and simulationsof larval dispersal suggest that larvae from Johnston atollcan reach French Frigate Shoals or Kaua‘i along two separatelarval corridors [49, 50].

The lollyfish, Holothuria atra, is one of the most commonshallow-water tropical sea cucumbers in the Indo-Pacific,spanning from Madagascar to French Polynesia [51, 52].H. atra performs vital ecosystem services on coral reefs forwhich there is an active fishery in many regions of thePacific [7, 53, 54]. Echinoderms are described as a boom-bustphylum in which populations go through marked naturalpopulation cycles [4], an attribute that can compound prob-lems in a harvested population but may hasten repopulationin previously impacted areas. As such, there is a call forecosystem-based management of sea cucumber harvests [7].Furthermore, the boom-bust nature of echinoderms hasimportant implications for connectivity in evolutionary timeframes, where biological attributes can drive populationstructure to a greater extent than oceanographic processesas hypothesized in the Tripneustes sea urchins [16]. Togetherthese characteristics make H. atra an ideal organism toexamine levels of connectivity and historical populationdynamics to inform management and to test hypothesesabout Hawai‘i’s connection with other archipelagos in theCentral Pacific. Here, we assess the inferred range of dispersalfor H. atra in Hawai‘i and the Central Pacific by investigatingits mitochondrial genetic population structure in an attemptto delineate the appropriate scales for management.

2. Methods

2.1. Sampling, PCR, and Sequencing. Holothuria atra wassampled from five archipelagos (Hawaiian Islands, LineIslands, Marshal Islands, Bonin Islands, and Ryukyu Islands)at a total of 19 sites (Figure 1). Sampling in the NorthwestHawaiian Islands and the Line Islands took place on researchcruises aboard the NOAA R.V. Hi‘ialakai. All other sampleswere collected on shore dives or while snorkeling. Samplingtook place between spring 2006 and fall 2009. Sampleswere obtained nonlethally through muscle-tissue biopsy andpreserved in either 95% ethanol or DMSO salt buffer andarchived at the Hawai‘i Institute of Marine Biology at roomtemperature. Skillings and Toonen [55] made an extendeddiscussion of sampling and preservation protocol. No asexualmorphs—distinguished by transverse scarring, smaller bodysize, and their location in lagoonal habitats—were foundduring sampling expeditions, and no reports are knownindicating the presence of the asexual stage of H. atra in thesampled locations. The asexual morph of H. atra appearsto be located only in the Southern and West Pacific (e.g.,[52, 56, 57]).

Total genomic DNA was extracted using DNeasy Bloodand Tissue Kits (QIAGEN) following the manufacturer’sinstructions. Polymerase chain reaction (PCR) was usedto amplify a 423-base pair fragment of the mitochondrialcytochrome c oxidase subunit I gene (COI) using customprimers created with Primer3 [58] targeting Holothuria spp.:GenHol2L (5′-AACCAAATGGTTCTTGCTTACC-3′) andGenHol2R (5′-TTCTGATTAATCCCACCATCC-3′). PCRwas performed using 15 μL reactions containing 1 μL ofdiluted DNA extract (one part template DNA to 199 partsnanopure water), 1 μL each of 0.2 μM forward and reverseprimers, 0.6 μL of 0.5 μM BSA, 7.5 μL of (Bioline) BiomixRed diluted as per manufacturer’s instructions, and 3.9 μLof nanopure water. PCR was done on Icycler thermocyclers(Bio-Rad Laboratories) with an initial denaturation at95◦C for 7 min followed by 35 cycles of a denaturingstep at 95◦C for 1 min, annealing at 50◦C for 1 min, andextension at 72◦C for 1 min. A final extension at 72◦C washeld for 7 min before refrigeration. PCR product (8 μL)was treated with 0.7 μL of Exonuclease I combined with0.7 μL of calf intestinal alkaline phosphatase (Exo-CIAP)and incubated at 37◦C for 30 minutes and with a finalinactivation step at 85◦C for 10 minutes. The treated PCRproduct was sequenced using an ABI Prism automaticsequencer at the Hawai‘i Institute of Marine Biology’sEPSCoR sequencing facility. All samples were sequenced inthe forward direction; uncertain sequences and all uniquehaplotypes were also sequenced in the reverse direction forconfirmation. Sequences were compiled and trimmed usingSequencher 4.8 and aligned using ClustalW implemented inBioedit 7.0.5 [59, 60].

2.2. Data Analysis. A statistical parsimony network ofmitochondrial haplotypes was constructed by creating areduced median network that was then used to make amedian joining network; both procedures implemented inNetwork 4.516 (http://www.fluxus-engineering.com/; [61,

Journal of Marine Biology 3

62]). The network was drawn using Network Publisher1.1.0.7 (http://www.fluxus-engineering.com/).

Nei’s average pairwise genetic difference (π) [63] andhaplotype diversity (h) were calculated in DnaSP 4.1 [64].The effective number of alleles was calculated by handfollowing Jost [65]. Tajima’s D [66] and Fu’s FS [67] werecalculated in Arlequin v.3.1 with 10,000 simulations toestablish confidence intervals.

To assess levels of genetic differentiation between sites,we calculated pairwise ΦST values using Arlequin 3.1 [68]and pairwise Dest chao values by hand following Jost [65].ΦST is a fixation index incorporating genetic distancethat ranges from 0 to 1, where a zero indicates identicalhaplotypic composition and a one signifies alternate fixationof alleles and a complete lack of gene flow. Dest chao is anindex of genetic differentiation, which does not account forgenetic distance among haplotypes but also ranges from0 to 1 (note that both ΦST and Dest chao can be slightlynegative due to bias correction for sampling error). In thecase of Dest chao, a zero also indicates identical haplotypiccomposition, but unlike ΦST, a one simply indicates thatno haplotypes are shared between the populations. Theprimary difference in interpretation is that in the absenceof gene flow ΦST values can be significantly less than one,while this is not the case for Dest chao, which is arguedto be an advantage of this latter statistic [65]. To correctthe critical P value for statistical significance in pairwisecomparisons, the familywise false discovery rate (FDR)correction found in [69] was implemented. Analysis ofmolecular variance (AMOVA) was used for hierarchicalanalysis of the partitioning of COI diversity among siteswithin archipelagic regions and among archipelagic regionsusing Arlequin 3.1. SAMOVA 1.0 was used to identify groupsof samples that maximize the proportion of total geneticvariance due to differences between regions [70]. The mostimportant genetic barriers were ranked using BARRIER2.2 [71]. BARRIER uses Monmonier’s maximum-differencealgorithm to compare a matrix of difference values, such aspairwise ΦST values, with a matrix of geographic distancesin order to identify the strongest barriers within the matrix.We compared barriers created using each ΦST and Dest chao

distance matrix. AMOVAs were performed using groupingsdetermined by SAMOVA and BARRIER for hypothesistesting to compare genetic groupings to the archipelagicgroupings. The pairwise ΦST and AMOVA analyses wereconducted using a distance matrix with 50,000 permutationsand the Tamura-Nei mutational model [72] with gamma =0.0164. The mutational model HKY+G was selected usingAIC in MODELTEST 3.7; the model hierarchy was used toselect the closest available model when the best-fit modelcould not be implemented by the chosen program, as inthe case of ARLEQUIN [73]. Regardless, the inferences arerobust to the mutational model, and our conclusions arenot altered regardless of which model is chosen (data notshown).

MRBAYES 3.1 was used to construct a Bayesian esti-mation of a phylogeny containing all H. atra haplotypesfrom this study along with all Holothuria and ActinopygaCOI haplotypes available from GenBank as of February

2010 [74]. Two independent runs with identical conditionswere completed and averaged. A general time reversible(GTR) simple nucleotide model with a gamma-shaped ratevariation of 0.0164 was used; Markov chain length = 4 ×3,000,000 sampled every 100 generations with a 10% burn-in. The GTR nucleotide model was chosen as it is the mostgeneral and neutral nucleotide model available in MrBayes3.1 and corresponds most closely to the Tamura-Nei model[75]. The sea cucumber Actinopyga agassizi was set as theoutgroup. Program defaults were used for all other settings.MrBayes was used to summarize all of the trees producedinto a single consensus tree.

RAxML 7.0 [76] implemented through CIPRES WebPortal v.1.15 [77] was used to construct the highest scor-ing maximum likelihood-based estimation of a phylogenycontaining all haplotypes used in the Bayesian analysis andrun 10,000 bootstrap simulations to assess branch support.A GTR nucleotide model that uses four discrete gammarates set by the program was used for the analysis; programdefaults were used for all other settings.

Bayesian coalescent-based calculations of migration rateamong regions (NeM) and the region mutation parameter(θ) were conducted using MIGRATE 3.1.3 [78]. Threeindependent runs of a Bayesian MCMC search strategy werecompleted and averaged by MIGRATE. A nucleotide modelwith a transition-to-transversion ratio of 6.1584 : 1 and threeregions of substitution rates with a gamma-shaped ratevariation of 0.016 was used; Markov chain length= 1,000,000sampled every 20 generations with a 10% burn-in. Programdefaults were used for all other settings. The transition-to-transversion ratio was calculated using Modeltest 3.7. Tworeplicate MIGRATE analyses were run using different pop-ulation groupings. Preliminary analyses that split the databy sampling location returned flat posterior probabilities,presumably from having too many parameters to estimate.The software’s author advocates using the minimal numberof sensible regions in order to reach convergence (PeterBeerli, pers. comm.). The first analysis used regions separatedalong the most important breaks identified by the programBARRIER. Archipelagos were used as regions for the secondanalysis with the Hawaiian Archipelago further dividedbetween the main Hawaiian Islands and the NorthwesternHawaiian Islands plus Johnston Atoll and Kingman Reefseparated from the Line Islands; this grouping was chosenbased on BARRIER and pairwise analyses. Values for themigration rate among regions (M) and region mutationparameters (θ) were taken from the highest peaks inthe posterior probability distribution curves. The posteriorprobability distributions were examined to determine thevalidity of each estimated parameter.

3. Results

A total of 385 individuals, 55 haplotypes, and 37 privatehaplotypes are sampled in this study (Table 1). Of the 18haplotypes shared across sampling sites, 10 are found inmultiple archipelagos (Figure 2). However, no haplotypes areshared between the most distant regions: Hawai‘i and Japan.


Okinawa

1

2

3

4

5

Japan

Philippines

New Guinea

MHI

Line Islands

NWHI

ChinaNiihau

Hilo

Jarvis

Palmyra

KonaOahu

Kauai

Marshall Is.

Ogasawara

Laysan

Pearl & Hermes

Johnston

Kingman

French Frigate

MidwayKure

Figure 1: Map of the northern Central and West Pacific. Pie charts represent haplotype frequencies overlaid on sites. To avoid overlap,the Marshall Islands sites, Kwajalein and Majuro, have been combined. Black lines are major gene flow constraints drawn by BARRIER,numbered from strongest to weakest. Yellow arrows represent directional migration rates estimated by MIGRATE. Effective migration rateestimates (NeM) with modes between 0.01 and 0.49 are represented by small arrows, rates between 0.5 and 0.99 are represented by mediumarrows, and rates over 1.0 are represented by large arrows. Solid lines represent migrate posterior probability distributions, where the 50%credibility set does not include zero. Dashed lines represent migration posterior probability distributions, where the 50% credibility setincludes zero, but the distribution peak is greater than zero. Migration parameters with a posterior distribution peak of zero are not shown.

Because many population genetic estimates are relativelyinsensitive to weak selection [79], loci which do not show sig-nificant deviations from neutral expectations should providereliable inferences about population structure [80]. Noneof the site-by-site Tajima’s D values were significant, andonly Laysan deviated from expectation using Fu’s Fs; thus,there is no evidence to indicate that nonneutral processesare responsible for the pattern of COI haplotype diversitypresented here.

To ensure there was no misidentification of the samplesincluded here, we performed a phylogenetic reconstructionof our samples with those available in GenBank. We con-firmed that the samples included here are monophyletic andthere are no reciprocally monophyletic groups among the H.atra haplotypes included in our population genetic analyses(Appendix A).

Haplotype diversity, as a function of longitude, increasesfrom west to east across the Pacific Ocean (Table 1, R2 =0.80, P < 0.05). Japanese sites exhibit the lowest haplotypediversity (h = 0.51–0.66) and effective number of haplotypes(HE = 2.0–2.9); the centrally located Line Islands (h = 0.65–0.90, HE = 2.9–9.7) and Marshall Islands (h = 0.75–0.76,HE = 4.0–4.1) exhibit mid to high levels of diversity; the Ha-waiian sites exhibit the highest diversity (h=0.75–0.90, HE =4–10). Nucleotide diversity does not appear to be correlatedwith haplotype diversity because nucleotide diversity is lowerin the Hawaiian Archipelago than in all other locationsexcept for Okinawa (Table 1). This pattern can be visualizedin the haplotype network, where sites exhibiting highnucleotide diversity harbor disparate haplotypes separatedby a relatively large number of mutations (Figure 2).

Four AMOVAs were run on the H. atra COI haplotypedata (Table 2). In each AMOVA, a different method wasemployed to group the population samples into regions.Under the first geographic hypothesis, sites were groupedinto five regions by archipelago, with Johnston Atoll includedwith Hawai‘i. In order to assess the subdivision of theHawaiian Archipelago into the NWHI and the MHI, thisgrouping hypothesis was compared to a second geographichypothesis, where population samples were grouped intosix regions; Hawaiian sites were divided into the MHIand the NWHI + Johnston and four regions comprised ofthe four remaining archipelagos. Both grouping hypotheses(geographic hypothesis one and two) exhibited similar parti-tioning of variation among groups (32.1% versus 30.9%) andamong populations within groups (8.9% versus 5.5%). Inboth cases, there was stronger partitioning among the groupsof samples (ΦCT = 0.32, ΦCT = 0.31; P < 0.0001) thanamong the simples nested within the groupings (ΦSC = 0.13,ΦSC = 0.07; P < 0.0001).

The two geographic hypotheses were compared to six-region groupings identified by SAMOVA and BARRIER;BARRIER selected the same dominant barriers using boththeΦST and Dest chao distance matrices (Table 2 and Figure 1).Six regions were chosen for a direct comparison to thearchipelagic geographic hypothesis that included Hawai‘idivided into two main regions. These groupings partitionedvariance similarly to the geographic hypotheses with theSAMOVA grouping minimizing among-population within-group variance with more among-group variance explained(Table 2). The grouping of population samples using BAR-RIER and SAMOVA had slightly greater levels of genetic


Johnston Atoll

Main Hawaiian Islands

Hilo, Hawaii

Kona, Hawaii

Oahu

Kauai

Niihau

Northwest Hawaiian Islands

French Frigate Shoals

Gardner Pinnacles

Laysan Island

Pearl and Hermes Atoll

Midway Atoll

Kure Atoll

Line Islands

Kingman Reef

Palmyra

Jarvis

Marshall Islands

Majuro

Kwajalein

Japan

Ogasawara

Okinawa

Southwest Pacific

Maluku

Papua New Guinea

Great Barrier Reef

New Caledonia

Unknown

8

44

Figure 2: Haplotype network for Holothuria atra. Each circle represents a unique haplotype connected by a line to those that differ by onebase pair. Nodes on lines indicate a missing haplotype, and numbers represent multiple missing haplotypes. Each haplotype is color codedby site, and circle size is proportional to frequency. The smallest circles represent one occurrence of a halotype.


Table 1: N is the Sample size, H is the total number of haplotypes, Hu is the number of unique haplotypes at site, π is the nucleotidediversity, h is the haplotype diversity, and AE is the effective number of alleles in COI.

Region Site N H Hu π ± SD h± SD AE Tajima’s D Fu’s Fs

Main Hawaiian Islands

Hilo 9 4 0 0.0041 ± 0.0030 0.81 ± 0.08 5.3 −0.27 0.08

Kona 21 10 2 0.0078 ± 0.0046 0.87 ± 0.06 7.5 −0.57 −2.21

Oahu 24 7 2 0.0052 ± 0.0033 0.79 ± 0.05 4.7 −0.88 −0.58

Kauai 30 8 2 0.0033 ± 0.0023 0.75 ± 0.06 4.0 −1.21 −2.61

Niihau 5 4 1 0.0071± 0.0052 0.90 ± 0.16 10.0 −0.75 −0.33

Northwest Hawaiian Islands

French Frigate 28 10 2 0.0082± 0.0048 0.88 ± 0.04 8.3 −0.14 −1.12

Gardner 2 1 0 N/A N/A N/A N/A N/A

Laysan 12 8 0 0.0064± 0.0041 0.89 ± 0.08 9.1 −1.06 − 2.91

Pearl & Hermes 37 10 2 0.0086± 0.0049 0.79 ± 0.05 4.8 −1.26 −0.23

Midway 35 14 6 0.0084 ± 0.0048 0.84 ± 0.05 6.2 −0.64 −3.61

Kure 23 8 2 0.0096 ± 0.0055 0.85 ± 0.04 6.9 −0.92 0.47

Johnston Johnston 26 7 1 0.0131± 0.0073 0.81 ± 0.05 5.3 −0.28 2.96

Line IslandsKingman 27 10 3 0.0142 ± 0.0078 0.90 ± 0.03 9.7 −0.39 0.81

Palmyra 16 4 2 0.0155 ± 0.0087 0.76 ± 0.06 4.1 2.59 6.13

Jarvis 23 5 3 0.0126 ± 0.0070 0.65 ± 0.08 2.9 1.8 4.84

Marshall IslandsMajuro 14 5 3 0.0139 ± 0.0079 0.76 ± 0.08 4.1 2.22 3.32

Kwajalein 9 4 1 0.0126 ± 0.0076 0.75 ± 0.11 4.0 1.49 2.81

Bonin Islands Ogasawara 20 4 3 0.013 ± 0.0073 0.66 ± 0.06 2.9 2.25 6.15

Ryukyu Islands Okinawa 24 3 2 0.0028 ± 0.0021 0.51 ± 0.09 2.0 0.26 1.82

Overall 385 55 37 0.0088 ± 0.0045 0.92 ± 0.08 12.5 0.12± 1.28 0.83± 2.91

differentiation among groups of samples (ΦCT = 0.33,ΦCT = 0.37; P < 0.0001) and lower levels of differentiationamong samples nested within groups (ΦSC = 0.07, ΦSC =0.03; P < 0.0001). Overall, the four AMOVAs exhibitedsimilar levels of partitioning of variance, and all tests weresignificant (P < 0.0001).

There is a strong pattern of restricted gene flow betweensites among the population samples of H. atra. Pairwisecomparisons for both ΦST and Dest chao reveal significantdifferences between sites located in different archipelagosin almost all cases, where sample sizes are 10 or greater(Table 3). The exceptions are between the Line Islandsand the Marshall Islands, where one of six pairwise ΦST

comparisons is statistically significant; between the LineIslands and the Bonin Islands, one of three ΦST comparisonsis statistically significant; between the Hawaiian Islands andKingman Reef, where only four of twelve pairwise ΦST

comparisons were statistically significant.Some significant differences were also detected among

samples within archipelagos. In the Main Hawaiian Islands(MHI), O‘ahu and Kaua‘i are significantly different than theKona sample from the Big Island of Hawai‘i. Despite smallsample sizes, Ni‘ihau is also partitioned from the adjacentisland of Kaua‘i as well as O‘ahu, but not the Big Island.Within the Northwestern Hawaiian Islands (NWHI), onlyLaysan is significantly partitioned from the other samplingsites, including Johnston Atoll. Overall, 42% of pairwisecomparisons between the MHI and the NWHI + Johnstonwere significant, compared to 30% of the comparisons withinthe MHI and 19% of the comparisons within the NWHI

+ Johnston. The samples from both the Line Islands (LI)and the Marshall Islands (MI) were significantly partitionedwithin their respective archipelagos when using Dest chao;there was not significant partitioning between MI sites whenusing ΦST.

The results from the MIGRATE runs show similarpatterns of gene flow between regions (Table 4, Figure 1).Effective migration rates (NeM) between regions are low.There is less than one migrant per generation, the rule-of-thumb number below which population cohesion startsto break down, between most regions [81]. The exceptionsinclude the one-way migration from the Main HawaiianIslands into the Northwest Hawaiian Islands and JohnstonAtoll and the one-way migration from Kingman Reef tothe Main Hawaiian Islands (Table 4). The high effectivemigration rate from the NWHI to the MHI in the firstanalysis splits almost evenly between Kingman Reef and theMHI when these two regions are separated in the secondanalysis (Table 4). Overall, higher effective migration ratesare observed leaving the Hawaiian regions then going intothem (Table 4, Figure 1). Though effective migration ratesare a product of migration and effective population size, theeffective migration rates larger than 1 migrant per generationare driven primarily by migration and not effective popu-lation size (Appendices A and B). This pattern is indicativeof recent migration rather than ancestral polymorphismsand high effective population sizes. Posterior probabilitydistributions for all values were in the form of unimodalcurves. A full description of Ne and M values for bothanalyses can be found in Appendices B and C.


Ta

ble

2:Pa

irw

ise

com

pari

son

sby

site

.Des

tch

aova

lues

are

con

tain

edin

the

low

erle

fth

alfo

fth

eta

ble,

andΦ

stva

lues

are

inth

eu

pper

righ

th

alfo

fth

eta

ble.

Bol

ded

valu

essi

gnif

ysi

gnifi

can

tdi

ffer

ence

saf

ter

corr

ecti

onu

sin

gth

epr

oced

ure

outl

ined

inB

enja

min

i200

8.

Reg

ion

Site

Mai

nH

awai

ian

Isla

nd

sN

orth

wes

tH

awai

ian

Isla

nd

sJo

hn

ston

Lin

eIs

lan

ds

Mar

shal

lIsl

and

sB

onin

Iss.

Ryu

kyu

Iss.

Hilo

Kon

aO

ahu

Kau

aiN

iihau

Fr.

Frig

ate

Gar

dn

erL

aysa

nPe

arl

&H

erm

esM

idw

ayK

ure

Kin

gman

Palm

yra

Jarv

isM

aju

roK

waj

alei

nO

gasa

war

aO

kin

awa

Mai

nH

awai

ian

Isla

nd

s

Hilo

—0.

101

−0.0

39−0

.029

0.23

40.

209

0.46

3−0

.059

0.25

30.

243

0.18

50.

204

0.08

40.

303

0.39

90.

416

0.15

90.

335

0.86

4

Kon

a0.

225

—0.

127

0.20

3−0

.047

−0.0

06−0

.044

0.04

90.

026

0.01

60.

009

0.04

60.

056

0.32

70.

401

0.36

10.

203

0.34

50.

786

Oah

u−0

.167

0.51

5—

0.04

90.

261

0.22

10.

399

0.01

80.

258

0.26

30.

193

0.23

10.

117

0.35

10.

431

0.46

80.

181

0.36

80.

826

Kau

ai−0

.109

0.49

90.

172

—0.

377

0.30

50.

568

0.00

10.

341

0.33

50.

278

0.30

90.

183

0.45

60.

524

0.57

30.

333

0.47

30.

873

Niih

au0.

245

−0.3

880.

516

0.69

8—

−0.0

11−0

.224

0.09

10.

006

−0.0

45−0

.005

0.00

70.

024

0.26

60.

359

0.26

90.

178

0.30

10.

847

Nor

thw

est

Haw

aiia

nIs

lan

ds

Fr.F

riga

te0.

392

−0.1

050.

518

0.62

9−0

.239

—−0

.039

0.14

5−0

.014

−0.0

120.

001

0.01

50.

084

0.35

80.

421

0.36

70.

256

0.37

10.

767

Gar

dn

er0.

814

0.41

21

10.

273

0.55

3—

0.23

924

−0.0

21−0

.091

0.03

3−0

.045

0.01

10.

258

0.37

50.

215

0.19

90.

315

0.88

9

Lay

san

−0.2

940.

138

0.12

2−0

.13

0.19

10.

286

0.84

9—

0.19

40.

173

0.12

40.

156

0.06

80.

306

0.39

60.

398

0.16

70.

334

0.83

2

Pear

l&

Her

mes

0.68

90.

008

0.74

40.

888

−0.1

1−0

.033

0.42

0.67

4—

−0.0

12−0

.013

−0.0

050.

089

0.36

50.

419

0.35

80.

269

0.37

20.

739

Mid

way

0.66

1−0

.08

0.84

60.

88−0

.356

−0.0

260.

360.

555

−0.0

18—

0.00

40.

0177

0.09

10.

365

0.42

10.

349

0.27

10.

373

0.74

7

Ku

re0.

461

0.11

10.

410.

666

−0.2

01−0

.03

0.69

70.

483

0.03

490.

16—

−0.0

130.

042

0.30

40.

369

0.32

90.

197

0.31

60.

739

Joh

nst

on0.

799

0.14

40.

782

0.90

10.

0909

−0.0

120.

611

0.69

4−0

.041

0.09

3−0

.022

—0.

071

0.29

70.

358

0.29

40.

205

0.30

80.

682

Lin

eIs

lan

ds

Kin

gman

0.27

90.

432

0.46

70.

354

0.78

10.

463

0.93

30.

142

0.64

30.

695

0.61

20.

666

—0.

145

0.18

30.

153

0.01

90.

152

0.57

1

Palm

yra

11

11

11

11

11

11

0.66

4−

0.06

50.

109

0.03

10.

009

0.45

1

Jarv

is0.

626

0.88

0.59

10.

730.

728

0.86

10.

776

0.94

10.

966

0.73

20.

956

0.52

10.

447

—0.

041

0.04

60.

064

0.28

9

Mar

shal

lIsl

and

sM

aju

ro1

11

11

11

11

11

10.

539

0.39

10.

241

—0.

099

0.10

70.

386

Kw

ajal

ein

0.22

20.

752

0.15

50.

499

0.49

20.

721

0.53

20.

883

0.93

80.

512

0.92

20.

510.

576

−0.0

340.

419

—0.

044

0.61

9

Bon

inIs

lan

ds

Oga

saw

ara

11

11

11

11

11

11

10.

743

10.

902

1—

0.42

8

Ryu

kyu

Isla

nd

sO

kin

awa

11

11

11

11

11

11

0.58

50.

433

0.17

20.

222

0.40

21

—

Fam

ilyw

ise

fals

edi

scov

ery

rate

corr

ecte

dP

-val

ues

:Des

tch

aoP≤.0

36;Φ

stP≤.0

23.


Table 3: Analysis of molecular variance (AMOVA) for different population groupings. Beginning at the top, groupings are as follows: (1)by archipelago, (2) by archipelago with Hawaii divided between the Northwest Hawaiian Islands (NWHI) and the Main Hawaiian Islands(MHI), (3) six groups chosen by BARRIER and (4) six groups chosen by SAMOVA.

Population groupings Source of variation % of Variation Φ statistics

Okinawa; Ogasawara; Marshall AG 32.14 ΦCT = 0.321∗

Islands; Hawaiian Archipelago + AP(G) 8.9 ΦSC = 0.131∗

Johnston Atoll; Line Islands WP 58.96


Islands; MHI; NWHI + Johnston Atoll; AP(G) 5.47 ΦSC = 0.0793∗

Line Islands WP 63.58


Islands; MHI + Kingman; NWHI + AP(G) 4.48 ΦSC = 0.066∗

Johnston Atoll; Palmyra + Jarvis WP 62.93

Okinawa; Ogasawara + Kwajalein + AG 37.35 ΦCT = 0.373∗

Palmyra; Jarvis + Majuro; Hilo + AP(G) 2.02 ΦSC = 0.032∗

Oahu + Kauai; All Others WP 60.63

AG = Among groups, AP(G) = Among populations within Groups, WP = Within populations. ∗P < 0.0001.

Table 4: Pairwise population migration rate estimates (NeM) based on a Bayesian MCMC simulation. Kingman Reef is grouped with theMain Hawaiian Islands in the upper table and treated as a seprate population in the lower table. The value of M calculated by MIGRATEwas multiplied by the θ, as calculated by MIGRATE, of the destination population to estimate migration. The estimates of migration areseparated by direction; the columns are source populations, and the the rows are sink populations.

Region Line Islands MHI+King NWHI+J Okinawa Ogasawara Marshall Islands

Line Islands — 0.13125 0.00375 0.15375 0.16875 0.13875

MHI+Kingman 0.02375 — 3.01625 0.02375 0.02375 0.45125

NWHI+Johnston 0.03125 4.34375 — 0.03125 0.03125 0.03125

Okinawa 0.04625 0.00875 0.02875 — 0.04375 0.04375

Ogasawara 0.04625 0.01375 0.00125 0.04875 — 0.02875

Marshall Islands 0.46125 0.55125 0.19125 0.43875 0.39375 —

Line Islands MHI NWHI+J Okinawa Ogasawara Marshall Islands Kingman

Line Islands — 0.04375 0.01125 0.05125 0.05125 0.04875 0.03125

MH Iss. 0.01625 — 0.95875 0.01625 0.27625 0.30875 1.02375

NWHI+Johnston 0.02625 4.96125 — 0.02625 0.02625 0.02625 0.49875

Okinawa 0.04625 0.05125 0.02875 — 0.04375 0.04875 0.04125

Ogasawara 0.04375 0.02625 0.03125 0.04375 — 0.04875 0.02875

Marshall Islands 0.46125 0.43875 0.41625 0.46125 0.39375 — 0.41625

Kingman Reef 0.69375 0.84375 0.80625 0.88125 0.01875 0.73125 —

4. Discussion

In this survey of population genetic structure, we elucidatepatterns of connectivity throughout the north-central rangeof the sea cucumber Holothuria atra with a focus on theHawaiian Archipelago. The Hawaiian Archipelago is highlyisolated and also contains one of the highest proportions ofendemism in the world (e.g., [40–42]; reviewed by Ziegler[39], Eldredge and Evenhuis [43]). Though there are manypan-pacific marine organisms in Hawai‘i, the isolation ofthe archipelago is thought to limit larval exchange such thatcolonization is rare but sufficient to maintain species cohe-sion among these taxa. The Hawaiian marine fauna containsa large proportion of endemics that are differentiated butnot diversified from its Indo-West Pacific roots [39, 46–48]. In this scenario, Hawai‘i is seen primarily as a dead

end, an isolated land mass that does not contribute in asignificant way to the overall diversity of the tropical pacific.Counter to the island biogeography hypotheses of Hawaiiandiversity, Jokiel and Martinelli [44] proposed the Vortexmodel of speciation, wherein the stunning biodiversity ofthe Coral Triangle is a result of centrifugal accumulationof species from the peripheral habitats around the Pacific.Though these two models primarily make predictions aboutspeciation-level processes and do not speak directly to geneflow within a species, they do make opposite claims about thedominant direction of gene flow and dispersal. H. atra hasa broad species range, extending from the Western IndianOcean to the Eastern Pacific Ocean, which suggests thecapacity for long-distance dispersal; however, populationsshowed significant population structuring within archipela-gos, sometimes across very short oceanic distances. Even so,


hierarchical genetic population structure in H. atra givesinsight into the phylogeography of the north-central tropicalPacific. Our data test between the divergent hypotheses ofwhether peripheral archipelagos act as a source of geneticdiversity in the Pacific and the likely colonization routes, intoand out of, the extremely isolated Hawaiian Archipelago.

4.1. Biogeography and Range Size. If a large species rangeis a consequence of high dispersal potential, then H.atra should have little pronounced population structure,especially across small scales [11, 15, 82]. Indeed, this is thecase for many species in the central West Pacific [16, 83–86].Despite a species range which stretches from the Western RedSea to the eastern Central Pacific in which H. atra is found inalmost all shallow tropical habitats, we did not find supportfor extensive dispersal. The majority of sites from which wesampled H. atra were genetically distinct, with some sites lessthan 75 km apart being among the most distinct in our study(Table 3). These contrasting patterns highlight the dangersof making predictions about population connectivity anddiversity based solely on the location and size of a species’range.

The larval life history of H. atra is not known exactly,but they require at least 18–25 days to reach competency tosettle and are capable of traversing long oceanic distanceswith sufficient frequency to maintain species cohesion acrossa very broad geographic range [87]. The obvious questionbecomes why then is population subdivision found on suchsmall geographic scales (e.g., Kingman Reef and PalmyraAtoll are only 67 km apart)? Counter to intuition, thegeographic distance among sites is a poor predictor of theease with which larvae can disperse among locations; the“oceanographic distance” experienced by larvae betweensites is uncorrelated with geographic separation betweenthem [36, 38]. Likewise, recent meta-analyses indicate therelationship between the length of pelagic larval develop-ment and dispersal ability is not as tight as has been generallyassumed [23–26, 88]. Finally, a broad meta-analysis by Lesteret al. [21] indicates that the intuitive relationship betweenrange size and larval dispersal potential is poorly correlatedoverall but can play an important role in some taxa. Toonenet al. in this issue also show a number of breaks in theHawaiian Archipelago that are shared by several species andare unexplained solely by appeal to one metric such as rangesize or larval dispersal potential. Although the mechanismof isolation across small scales remains unknown, our dataclearly indicate that H. atra is not one of those species forwhich range size predicts relative dispersal ability.

4.2. Population Structure in the Hawaiian Archipelago andJohnston Atoll. Our mtDNA examination of Holothuria atrareveals significant genetic population structure across thesurveyed portion of the range. There are two interestingpatterns to this structure. Excluding Laysan Island, there areno significant pairwise differences between any other islandsin the NWHI (spanning nearly 2000 km), suggesting thatthe NWHI, excluding Laysan, comprises a single large pop-ulation. In contrast, there is significant structuring within

the MHI (roughly 600 km) and between the NWHI andthe MHI. This finding suggests that factors beyond merelygeographic distance influence population partitioning.

Johnston Atoll, the nearest neighboring land mass,roughly 860 km south of French Frigate Shoals, is geneticallydistinct from most of the MHI and Laysan and geneticallysimilar to all of the NWHI except Laysan. It has beensuggested that Johnston Atoll acts as a stepping stone intothe Hawaiian Islands [89]. Kobayashi [49, 50] used computersimulations to predict two larval transport corridors fromJohnston Atoll to the Hawaiian Archipelago: one corridorstretching from Johnston to French Frigate Shoals in theNWHI and one from Johnston to O‘ahu in the MHI.Our data support the predicted larval transport corridorbetween Johnston Atoll and French Frigate Shoals, butnot the corridor predicted between Kaua‘i and Johnston.Additionally, based on our data, Kingman Reef may also bean important stepping stone into and out of Hawai‘i. TheBARRIER analysis shows the division between the NWHIs,including Johnston Atoll, and the MHI to be the strongestbarrier to gene flow within the Archipelago (Figure 1).Migration across this barrier is heavily one sided, wheremigration from the MHI into the NWHI dominates. Theeffectively one-way migration rates into the NWHI andJohnston Atoll coupled with the strong genetic similaritybetween Johnston Atoll and the NWHI suggest JohnstonAtoll is an isolated outpost of the Northwest HawaiianIslands, providing support for a vortex model [44] ratherthan the stepping stone entry into Hawai‘i [89] for H. atra.These data indicate that Johnston Atoll exchanges migrantswith Hawai‘i far more often than its nearest neighbors tothe south, and the same can be said for Kingman Reef.This result is particularly surprising because in the case ofKingman Reef, H. atra sampled there show greater similarityto populations in Hawai‘i (roughly 1700 km southwest ofHonolulu) than they do to those sampled at Palmyra Atoll,only 67 km away.

4.3. Phylogeographic Relationships between Archipelagos.Counter to conventional wisdom that Hawai‘i is a passiverecipient of rare dispersal from the diverse Pacific, the weightof available evidence, including pairwise ΦST values, mtDNAphylogeny, BARRIER divisions, and clustering within thehaplotype network, provides substantial evidence for theopposite pattern in H. atra; Johnston Atoll is an outpost ofHawaiian diversity, and Kingman Reef acts as the primarystepping stone between the Hawaiian Archipelago and therest of the Pacific. As far as we are aware, this is thefirst time empirical evidence has been provided for sucha pathway. Higher haplotypic diversity in Hawai‘i and theLine Islands relative to the other archipelagos supportsa scenario in which population sizes are far greater, orHawai‘i and/or Kingman Reef are the ancestral populationin the region. Likewise, the dominant haplotypes foundin the Japanese Archipelagos are relatively distantly relatedand appear derived (Figure 2), suggesting that the westernportion of the surveyed range was colonized in at least twoseparate events (or one of them has gone extinct in Hawai‘i


and Kingman), one of which did not make it all the way toOkinawa.

Excluding the dense sampling within the HawaiianArchipelago, virtually all pairwise comparisons between sitesare significantly different from each other (Table 3). The fewcomparisons that were not significant between sites haverelatively high pairwise values but low sample sizes, a likelystatistical limitation also noted by Bird et al. [90]. Hierarchi-cal population structuring was detected with AMOVA usingeither Dest or ΦST values; sampling sites within archipelagosare significantly different from each other but are moresimilar within than between archipelagos. Four of the fivemost substantial restrictions to gene flow uncovered in thisstudy (and the top ranked by BARRIER) were those betweenarchipelagos; the one exception to this trend being KingmanReef which is included with the Hawaiian rather than theLine Islands (Figure 1). The AMOVA run using the regionsselected by BARRIER was only minimally different than theAMOVA run using regions divided by archipelagos; groupingKingman Reef with the MHI, as per BARRIER, did explain1.5% more of the overall variance.

In addition to the distinct archipelagic groupings inthe haplotype network (Figure 2), several other patternsare noteworthy here. First, the NWHI and Johnston Atollhaplotypes are clustered together and interspersed whereasthe MHI haplotypes are clustered together. Also, the Japanesehaplotypes occur in two divergent areas of the network.The Line Island and Marshall Island haplotypes are inter-dispersed throughout the network, suggesting that theseisland groups are either mixing or transition zones. Nearlyevery locality haplotype (those found in only one samplinglocation) branch off in a starburst pattern from the majorhaplotypes found primarily in the same archipelago. Thispattern is an indication that regional populations havebeen separated long enough for new haplotypes to arise,and that these new haplotypes are not being spread toother archipelagos by long-distance dispersal. Uniformlylow migration rates between archipelagos estimated withMIGRATE support this isolation scenario.

5. Conclusion

Many echinoderm species are the focus of artisanal orcommercial fishing efforts, and managing these fisheriesrequires a detailed understanding of dispersal pathwaysand population connectivity within a spatial managementnetwork. The Hawaiian Archipelago lies at the periphery ofthe tropical Central Pacific and is the most isolated islandchain in the world; the question remains as to why somespecies maintain connectivity and species cohesion betweenthe Hawaiian Islands and the rest of the Pacific, why somespecies diverge and become Hawaiian endemics, and whyother species with similar inferred dispersal ability fail tocolonize the Hawaiian Archipelago at all.

The genetic diversity of COI in H. atra across thestudied portion of the range presents a complex pattern,but it is not inscrutable. Based on AMOVA, SAMOVA, andBARRIER analyses, it can be seen that population structuring

is hierarchical; there are significant differences between sites,but the primary degree of population structure is archipelagoby archipelago. Our analyses taken together suggest thatthe Hawaiian Archipelago and Kingman reef are ancestralpopulations in the region with migration moving out of theseperiphery archipelagos toward a less diverse central Pacificrather than the reverse. This pattern is inconsistent withthe hypothesis that Hawai‘i is a dead end for rare migrantsfrom the Indo-Pacific. Instead, the weight of the evidenceshows that these peripheral populations are not sinks, butimportant centers for the generation of genetic diversityfeeding back towards the West Pacific. Specifically for H. atra,our data suggest that the pathway between Hawai‘i and therest of the Pacific is primarily out through Kingman Reefand the Line Islands and not in through Japan, the MarshallIslands, or the closest neighbor to the Hawaiian Archipelago,Johnston Atoll. We show that, at least for H. atra, JohnstonAtoll is in fact an outpost of the Northwestern HawaiianIslands and not a primary gateway for colonization of theArchipelago.

Considerable evidence is accumulating that it is inde-fensible to make predictions of connectivity based solely onproxies such as ecological or phylogenetic similarity, pelagiclarval duration, or species range sizes [21, 23–25, 90]. Thefine-scale structuring of populations in H. atra suggeststhat place-based management approaches, as exemplifiedby ecosystem-based management, are ideal for respondingto the complex relationships between genetically distinctpopulations. Holothuria atra must be managed on a localscale; migration between archipelagos, and often betweenislands, does not occur in ecologically relevant time frames.

Appendices

A.

Phylogenetic tree of sampled Holothuria atra haplotypesand Holothuria and Actinopyga haplotypes retrieved fromGenBank. The maximum likelihood tree is shown; Bayesiananalysis produced a tree with almost identical topology.The first of the numbers beside branches corresponds tobootstrap support calculated from the maximum likelihoodanalysis; the second number, seperated by a /, correspondsto the posterior probability estimated during the Bayesiananalysis. Only support values with 50% majority rule or 0.50posterior probability are included. Actinopyga agassizi is theoutgroup. All non-H. atra nodes were collapsed. GenBankaccession numbers are included for uncollapsed taxa (seeFigure 3).

B.

M and θ posterior probability distributions were calculatedby MIGRATE using a Bayesian MCMC simulation. Popu-lation key: 1 = Kingman Reef; 2 = Line Islands; 3 = MainHawaiian Islands; 4 = Northwestern Hawaiian Islands andJohnston Atoll; 5 = Okinawa; 6 = Ogasawara; 7 = MarshallIslands (see Table 5).


Table 5

Parameter 2.50% 25.00% Mode 75.00% 97.50% Median Mean

θ1 0 0.0015 0.00375 0.0055 0.0115 0.00525 0.00518

θ2 0 0 0.00025 0.0015 0.0035 0.00175 0.00114

θ3 0 0.0015 0.00325 0.0045 0.007 0.00375 0.00345

θ4 0.001 0.0035 0.00525 0.007 0.0105 0.00625 0.00586

θ5 0 0 0.00025 0.0015 0.003 0.00175 0.00095

θ6 0 0 0.00025 0.001 0.003 0.00125 0.00085

θ7 0 0.0005 0.00225 0.0045 0.011 0.00375 0.00417

M2–>1 0 0 185 350 1150 345 426.5

M3–>1 0 90 225 560 1430 495 580.3

M4–>1 0 80 215 580 1460 525 609.3

M5–>1 0 110 235 530 1310 455 536.8

M6–>1 0 0 5 160 580 165 204.9

M7–>1 0 10 195 350 1110 335 417.3

M1–>2 0 0 125 260 1010 265 347.2

M3–>2 0 0 175 280 1000 285 363.1

M4–>2 0 0 45 220 860 225 293.6

M5–>2 0 20 205 490 1500 475 575.1

M6–>2 0 70 205 580 1500 525 612.6

M7–>2 0 0 195 360 1350 365 481.7

M1–>3 10 140 315 580 1490 515 609.7

M2–>3 0 0 5 160 620 165 208.8

M4–>3 0 150 295 660 1490 565 644.1

M5–>3 0 0 5 100 370 105 126.2

M6–>3 0 10 85 210 600 195 227.9

M7–>3 0 0 95 220 920 225 299.9

M1–>4 0 0 95 220 910 225 299.7

M2–>4 0 0 5 100 350 105 123.3

M3–>4 350 640 945 1340 1950 1085 1094.5

M5–>4 0 0 5 70 280 75 93.9

M6–>4 0 0 5 80 280 85 96.6

M7–>4 0 0 5 140 470 145 170.4

M1–>5 0 0 165 280 1050 285 371.1

M2–>5 0 0 185 350 1290 355 457.5

M3–>5 0 60 205 440 1260 385 480.6

M4–>5 0 0 115 260 960 265 344

M6–>5 0 0 175 300 1200 305 412.1

M7–>5 0 20 195 380 1200 355 444.4

M1–>6 0 0 115 270 1020 275 358

M2–>6 0 0 175 340 1240 345 440.7

M3–>6 0 0 105 250 910 255 322.7

M4–>6 0 0 125 250 950 255 330.6

M5–>6 0 0 175 310 1020 305 377.4

M7–>6 0 20 195 370 1170 345 434.6

M1–>7 0 0 185 320 1040 315 387.3

M2–>7 0 30 205 570 1580 535 637.2

M3–>7 0 40 195 440 1280 405 496.2

M4–>7 0 0 185 300 1040 305 385.1

M5–>7 0 0 205 580 1640 575 671.4

M6–>7 0 10 175 280 960 275 344


Table 6

Parameter 2.50% 25.00% Mode 75.00% 97.50% Median Mean

θ1 0 0 0.0008 0.0015 0.0035 0.00175 0.00125

θ2 0.001 0.003 0.0048 0.0065 0.01 0.00575 0.00532

θ3 0.002 0.0045 0.0063 0.0085 0.012 0.00725 0.00709

θ4 0 0 0.0003 0.0015 0.003 0.00175 0.00099

θ5 0 0 0.0003 0.0015 0.004 0.00175 0.00127

θ6 0 0.0005 0.0023 0.004 0.0105 0.00375 0.00399

M2–>1 0 0 175 270 1000 275 356

M3–>1 0 0 5 210 840 215 286.9

M4–>1 0 10 205 530 1570 515 616.5

M5–>1 0 80 225 550 1410 485 573.5

M6–>1 0 0 185 340 1260 345 447.1

M1–>2 0 0 5 210 750 215 265.9

M3–>2 160 390 635 950 1700 805 858.6

M4–>2 0 0 5 120 420 125 150.1

M5–>2 0 0 5 80 310 85 106.6

M6–>2 0 0 95 230 790 235 294.7

M1–>3 0 0 5 80 320 85 108.3

M2–>3 210 460 695 1030 1750 885 926.5

M4–>3 0 0 5 60 210 65 70

M5–>3 0 0 5 100 290 105 111.2

M6–>3 0 0 5 100 360 105 123.9

M1–>4 0 0 185 390 1440 395 511.3

M2–>4 0 0 35 230 860 235 298.3

M3–>4 0 0 115 240 880 245 315.2

M5–>4 0 0 175 320 1170 325 417

M6–>4 0 0 175 330 1280 335 447.2

M1–>5 0 0 185 330 1190 325 422.7

M2–>5 0 0 55 230 930 235 316

M3–>5 0 0 5 190 700 195 243.8

M4–>5 0 40 195 380 1160 345 430.6

M6–>5 0 0 115 250 940 255 328.9

M1–>6 0 10 205 550 1640 535 650.1

M2–>6 0 120 245 570 1410 505 584.5

M3–>6 0 0 85 230 880 235 307.3

M4–>6 0 0 195 470 1510 475 581

M5–>6 0 0 175 270 940 275 345.8


100/--

83/0.95

93/0.98

100/0.99

100/0.99

100/0.98

100/0.93100/0.93

46/0.93

67/0.82

33/0.59

33/0.77

100/0.9983/0.92

96/0.9697/0.99

47/0.88

100/0.84

54/0.88

100/0.96

65/0.89

100/0.8768/0.76

96/0.89

31/0.51

79/0.91

57/0.60

41/0.56

93/0.84

0.03

98/0.91

100/0.86100/0.98

96/0.87

20/0.65

29/0.5416/0.52

54/0.84

Actinopyga obesa

Actinopyga agassizi

Actinopyga mauritiana

Actinopyga sp.

EU848228

EU848231

EU848232

EU220815

EU848220

EU848219

FJ971394

Actinopyga caerulea

Holothuria spinea

Holothuria leucospilota

Holothuria coluber

Holothuria polii

Holothuria mexicana

Holothuria mexicana

Holothuria floridana

Holothuria signata

Holothuria edulis

Holothuria atra 02

Holothuria atra 41

Holothuria atra 22

Holothuria atra 53

Holothuria atra 01

Holothuria atra 20

Holothuria atra 39

Holothuria atra 38

Holothuria atra 19

Holothuria atra 12

Holothuria atra 36

Holothuria atra 37

Holothuria atra 07

Holothuria atra 10

Holothuria atra 13

Holothuria atra 43

Holothuria atra 03

Holothuria atra 04

Holothuria atra 09

Holothuria atra 08

Holothuria atra 06

Holothuria atra 11

Holothuria atra 24

Holothuria atra 17

Holothuria atra 16

Holothuria atra 52

Holothuria atra 42

Holothuria atra 55

Holothuria atra 49

Holothuria atra 45

Holothuria atra 50

Holothuria atra 46

Holothuria atra 44

Holothuria atra 51

Holothuria atra 48

Holothuria atra 47

Holothuria atra 54

Holothuria atra 33

Holothuria atra 28

Holothuria atra 13

Holothuria atra 31

Holothuria atra 23

Holothuria atra 26

Holothuria atra 30

Holothuria atra 34

Holothuria atra 25

Holothuria atra 35

Holothuria atra 29

Holothuria atra 32

Holothuria alra 27

Holothuria atra 15

Holothuria fuscogilva

Holothuria nobilis

Holothuria whitmaei

Holothuria echinites

Actinopyga palauensis

EU848296

FJ971397

EU220821

EU220822

AY700770

New Guinea

Cuba

Palau

Unknown

New Caledonia

New Guinea

Australia/New Caledonia


Unknown

Unknown

Unknown

Unknown

Unknown


Mediterranean

Australia, New Caledonia, Unknown

Kingman

MHI/NWHI/Kingman

MHI/NWHI

French Frigate

Midway

Pearl and Hermes

Great Barrier Reef

Midway

Kure

Johnston

NWHl/Johnston

Kingman

Ishigaki

Indonesia/Palmyra/Majuro/New Caledonia/Ogasawara/Papua New Guinea

Johnston

Majuro

Jarvis

LI/MI/lshigaki/Papua New Guinea

Ogasawara

Ogasawara

Ishigaki

Unknown

Kauai

Midway

NWHI/Johnston/MHI

Johnston/MHI/NWHI/Johnston/Kingman/Great Barrier Reef

Majuro

Pearl and Hermes

Midway

Majuro

Kure

MidwayNiihau

Kona-Big Island

Kingman

NWHI/Johnston/Unknown

MHI/NWHI

MHI/NWHI/Kingman/Kwajalein/Great Barrier Reef

Midway

New Caledonia

MHI

Oahu

Kwajalein

MHI/NWHI/Johnston/LI/MI

Jarvis

MHI/NWHI/Kingman

Oahu

MHI/NWHI

Kauai/Great Barrier Reef

Ogasawara

Palmyra

Australia/Fiji/Samoa/New Caledonia

Madagascar/La Reunion/Seychelles

Australia/Palau/New Caledonia/Indonesia

Figure 3

C.

M and θ posterior probability distributions were calculatedby MIGRATE using a Bayesian MCMC simulation. Popu-lation key: 1 = Line Islands; 2 = Main Hawaiian Islands +Kingman; 3 = Northwestern Hawaiian Islands + Johnston;4 = Okinawa; 5 = Ogasawara; 6 = Marshall Islands (seeTable 6).

Acknowledgments

The authors thank the Papahanaumokuakea MarineNational Monument, US Fish and Wildlife Services,and Hawai‘i Division of Aquatic Resources (DAR) forcoordinating research activities and permitting, and

the National Oceanic and Atmospheric Administration(NOAA) research vessel Hi‘ialakai and her crew for yearsof outstanding service and support. Special thanks go to B.Bowen, the members of the ToBo Lab, UH Dive Program,NMFS, PIFSC, CRED, M. Skillings, K. Boyle, J. Claisse, D.Wagner, P. Aldrich, M. Iacchei, J. Puritz, J. Eble, I. Baums, M.Timmers, N. Yasuda, R. Kosaki, S. Karl, C. Meyer, S. Godwin,M. Stat, X. Pochon, H. Kawelo, T. Daly-Engel, M. Craig,L. Rocha, M. Gaither, G. Conception, Y. Papastamatiou,M. Crepeau, Z. Szabo, J. Salerno, and the HIMB NSF-EPSCoR Core Genetics Facility. The authors also thank theanonymous reviewers who put in the extra time to helpstrengthen the quality of this work. This work was fundedin part by Grants from the National Science Foundation(nos. DEB#99-75287, OCE#04-54873, OCE#06-23678, and


OCE#09-29031), National Marine Sanctuaries NWHICRER-HIMB partnership (no. MOA-2005-008-6882), NationalMarine Fisheries Service, NOAA’s Coral Reef ConservationProgram, and the Hawai‘i Coral Reef Initiative. This iscontribution no. 1421 from the Hawai‘i Institute of MarineBiology and SOEST 8049.

References

[1] R. T. Paine, “A note on trophic complexity and communitystability,” The American Naturalist, vol. 103, no. 929, pp. 91–93, 1969.

[2] M. E. Power, D. Tilman, J. A. Estes et al., “Challenges in thequest for keystones: identifying keystone species is difficult-but essential to understanding how loss of species will affectecosystems,” BioScience, vol. 46, no. 8, pp. 609–620, 1996.

[3] H. A. Lessios, M. J. Garrido, and B. D. Kessing, “Demographichistory of Diadema antillarum, a keystone herbivore onCaribbean reefs,” Proceedings of the Royal Society B, vol. 268,no. 1483, pp. 2347–2353, 2001.

[4] S. Uthicke, B. Schaffelke, and M. Byrne, “A boom-bustphylum? Ecological and evolutionary consequences of densityvariations in echinoderms,” Ecological Monographs, vol. 79, no.1, pp. 3–24, 2009.

[5] S. A. Sloan, “Echinoderm fisheries of the world: a review,”in Echinodermata: Proceedings of the 5th International Echino-derm Conference, B. F. Keegan and B. D. S. O’Connor, Eds., pp.109–124, A.A. Balkema, Rotterdam, The Netherlands, 1984.

[6] E. Sala, C. F. Boudouresque, and M. Harmelin-Vivien, “Fish-ing, trophic cascades, and the structure of algal assemblages:evaluation of an old but untested paradigm,” Oikos, vol. 82,no. 3, pp. 425–439, 1998.

[7] S. W. Purcell, “Managing sea cucumber fisheries with anecosystem approach,” in FAO Fisheries and Aquaculture Tech-nical Paper No. 520, A. Lovatelli, M. Vasconcellos, and Y.Yimin, Eds., FAO, Rome, Italy, 2010.

[8] D. Hedgecock, P. H. Barber, and S. Edmands, “Geneticapproaches to measuring connectivity,” Oceanography, vol. 20,no. 3, pp. 70–79, 2007.

[9] S. R. Thorrold, D. C. Zacherl, and L. A. Levin, “Populationconnectivity and larval dispersal using geochemical signaturesin calcified structures,” Oceanography, vol. 20, no. 3, pp. 80–89,2007.

[10] M. J. Fogarty and L. W. Botsford, “Population connectivityand spatial management of marine fisheries,” Oceanography,vol. 20, no. 3, pp. 112–123, 2007.

[11] G. Thorson, “Reproductive and larval ecology of marinebottom invertebrates,” Biological Reviews of The CambridgePhilosophical Society, vol. 25, pp. 1–45, 1950.

[12] R. R. Strathmann, “Hypotheses on the origins of marinelarvae,” Annual Review of Ecology and Systematics, vol. 24, pp.89–117, 1993.

[13] B. P. Kinlan and S. D. Gaines, “Propagule dispersal in marineand terrestrial environments: a community perspective,” Ecol-ogy, vol. 84, no. 8, pp. 2007–2020, 2003.

[14] B. P. Kinlan, S. D. Gaines, and S. E. Lester, “Propagule dispersaland the scales of marine community process,” Diversity andDistributions, vol. 11, no. 2, pp. 139–148, 2005.

[15] G. Paulay and C. Meyer, “Dispersal and divergence acrossthe greatest ocean region: do larvae matter?” Integrative andComparative Biology, vol. 46, no. 3, pp. 269–281, 2006.

[16] H. A. Lessios, J. Kane, and D. R. Robertson, “Phylogeographyof the pantropical sea urchin Tripneustes: contrasting patterns

of population structure between oceans,” Evolution, vol. 57,no. 9, pp. 2026–2036, 2003.

[17] L. A. Levin, “Recent progress in understanding larval dispersal:new directions and digressions,” Integrative and ComparativeBiology, vol. 46, no. 3, pp. 282–297, 2006.

[18] R. K. Grosberg and C. W. Cunningham, “Genetic structurein the sea: from populations to communites,” in MarineCommunity Ecology, M. D. Bertness, S. D. Gaines, and M. E.Hay, Eds., pp. 61–84, Sinauer Associates, Sunderland, Mass,USA, 2001.

[19] K. A. Selkoe, C. M. Henzler, and S. D. Gaines, “Seascapegenetics and the spatial ecology of marine populations,” Fishand Fisheries, vol. 9, no. 4, pp. 363–377, 2008.

[20] M. E. Hellberg, “Gene flow and isolation among populationsof marine animals,” Annual Review of Ecology, Evolution, andSystematics, vol. 40, pp. 291–310, 2009.

[21] S. E. Lester, B. I. Ruttenberg, S. D. Gaines, and B. P. Kinlan,“The relationship between dispersal ability and geographicrange size,” Ecology Letters, vol. 10, no. 8, pp. 745–758, 2007.

[22] I. R. Bradbury and P. Bentzen, “Non-linear genetic isolation bydistance: implications for dispersal estimation in anadromousand marine fish populations,” Marine Ecology Progress Series,vol. 340, pp. 245–257, 2007.

[23] I. R. Bradbury, B. Laurel, P. V. R. Snelgrove, P. Bentzen,and S. E. Campana, “Global patterns in marine dispersalestimates: the influence of geography, taxonomic category andlife history,” Proceedings of the Royal Society B, vol. 275, no.1644, pp. 1803–1809, 2008.

[24] K. Weersing and R. J. Toonen, “Population genetics, larvaldispersal, and connectivity in marine systems,” Marine EcologyProgress Series, vol. 393, pp. 1–12, 2009.

[25] A. L. Shanks, “Pelagic larval duration and dispersal distancerevisited,” Biological Bulletin, vol. 216, no. 3, pp. 373–385,2009.

[26] P. M. Ross, I. D. Hogg, C. A. Pilditch, and C. J. Lundquist,“Phylogeography of New Zealand’s coastal benthos,” NewZealand Journal of Marine and Freshwater Research, vol. 43, pp.1009–1027, 2009.

[27] R. Burton and M. Feldman, “Population genetics of coastaland estuarine invertebrates: are behavioral differences amongspecies reflected in population structure?” Estuaries, vol. 4,article 239, 1981.

[28] N. Knowlton and B. Keller, “Larvae which fall short of theirpotential: highly localized recruitment in an alpheid shrimpwith extended larval development,” Bulletin of Marine Science,vol. 39, pp. 213–223, 1986.

[29] A. L. Shanks, B. A. Grantham, and M. H. Carr, “Propaguledispersal distance and the size and spacing of marine reserves,”Ecological Applications, vol. 13, no. 1, pp. S159–S169, 2003.

[30] E. G. Severance and S. A. Karl, “Contrasting populationgenetic structures of sympatric, mass-spawning Caribbeancorals,” Marine Biology, vol. 150, no. 1, pp. 57–68, 2006.

[31] M. N. Dawson and W. M. Hamner, “A biophysical perspectiveon dispersal and the geography of evolution in marine andterrestrial systems,” Journal of the Royal Society Interface, vol.5, no. 19, pp. 135–150, 2008.

[32] E. Bermingham and H. Lessios, “Rate variation of protein andmtDNA evolution as revealed by sea urchins separated by theIsthmus of Panama,” Proceedings of the National Academy ofSciences of the United States of America, vol. 90, pp. 2734–2738,1993.

[33] M. N. Dawson, “Phylogeography in coastal marine animals: asolution from California?” Journal of Biogeography, vol. 28, no.6, pp. 723–736, 2001.


[34] P. H. Barber, S. R. Palumbi, M. V. Erdmann, and M. K. Moosa,“Sharp genetic breaks among populations of Haptosquillapulchella (Stomatopoda) indicate limits to larval transport:patterns, causes, and consequences,” Molecular Ecology, vol.11, no. 4, pp. 659–674, 2002.

[35] E. E. Sotka, J. P. Wares, J. A. Barth, R. K. Grosberg, and S. R.Palumbi, “Strong genetic clines and geographical variation ingene flow in the rocky intertidal barnacle Balanus glandula,”Molecular Ecology, vol. 13, no. 8, pp. 2143–2156, 2004.

[36] I. B. Baums, C. B. Paris, and L. M. Cherubin, “A bio-oceanographic filter to larval dispersal in a reef-buildingcoral,” Limnology and Oceanography, vol. 51, no. 5, pp. 1969–1981, 2006.

[37] E. A. Treml, P. N. Halpin, D. L. Urban, and L. F. Pratson,“Modeling population connectivity by ocean currents, agraph-theoretic approach for marine conservation,” Land-scape Ecology, vol. 23, no. 1, pp. 19–36, 2008.

[38] C. White, K. A. Selkoe, J. Watson, D. A. Siegel, D. C. Zacherl,and R. J. Toonen, “Ocean currents help explain populationgenetic structure,” Proceedings of the Royal Society B, vol. 277,no. 1688, pp. 1685–1694, 2010.

[39] A. C. Ziegler, Hawaiian Natural History, Ecology and Evolution,University of Hawaii Press, Honolulu, Hawaii, USA, 2002.

[40] J. C. Briggs, Marine Zoogeography, McGraw-Hill, New York,NY, USA, 1974.

[41] E. A. Kay, “Little worlds of the Pacific. An essay on PacificBasin Biogeography,” Lecture 9, Harold L. Lyon Arboretum,University of Hawaii, pp. 1–40, 1980.

[42] R. W. Grigg, “Community structure, succession and develop-ment of coral reefs in Hawaii,” Marine Ecology Progress Series,vol. 11, pp. 1–14, 1983.

[43] L. G. Eldredge and N. L. Evenhuis, “Hawaii’s biodiversity: adetailed assessment of the numbers of species in the HawaiianIslands,” Bishop Museum Occasional Papers, vol. 76, pp. 1–28,2003.

[44] P. Jokiel and F. J. Martinelli, “The vortex model of coral reefbiogeography,” Journal of Biogeography, vol. 19, no. 4, pp. 449–458, 1992.

[45] E. A. Kay, “Patterns of speciation in the Indo-West Pacific,” inBiogeography of the Tropical Pacific, F. J. Radovsky, P. H. Raven,and S. H. Sohmer, Eds., pp. 15–31, Association of SystematicsCollections and B.P. Bishop Museum, 1984.

[46] E. A. Kay and S. R. Palumbi, “Endemism and evolutionin Hawaiian marine invertebrates,” Trends in Ecology andEvolution, vol. 2, no. 7, pp. 183–186, 1987.

[47] T. F. Hourigan and E. S. Reese, “Mid-ocean isolation andthe evolution of Hawaiian reef fishes,” Trends in Ecology andEvolution, vol. 2, no. 7, pp. 187–191, 1987.

[48] P. L. Jokiel, “Ecology, biogeography and evolution of corals inHawaii,” Trends in Ecology and Evolution, vol. 2, no. 7, pp. 179–182, 1987.

[49] D. R. Kobayashi, “Colonization of the Hawaiian Archipelagovia Johnston Atoll: a characterization of oceanographic trans-port corridors for pelagic larvae using computer simulation,”Coral Reefs, vol. 25, no. 3, pp. 407–417, 2006.

[50] D. R. Kobayashi and J. J. Polovina, “Simulated seasonal andinterannual variability in larval transport and oceanographyin the Northwestern Hawaiian Islands using satellite remotelysensed data and computer modeling,” Atoll Research Bulletin,no. 543, pp. 365–390, 2006.

[51] A. M. Clark and F. E. W. Rowe, Monograph of Shallow WaterIndo-West Pacific Echinoderms, Trustees of the British Museum(Natural History), London, UK, 1971.

[52] C. Conand, “Asexual reproduction by fission in Holothuriaatra: variability of some parameters in populations from thetropical Indo-Pacific,” Oceanologica Acta, vol. 19, no. 3-4, pp.209–216, 1996.

[53] K. Bonham and E. E. Held, “Ecological observations on thesea cucumbers Holothuria atra and H. leucospilota at RongelapAtoll, Marshall Islands,” Pacific Science, vol. 17, pp. 305–314,1963.

[54] S. Uthicke, “Sediment bioturbation and impact of feed-ing activity of Holothuria (Halodeima) atra and Stichopuschloronotus, two sediment feeding holothurians, at LizardIsland, great barrier reef,” Bulletin of Marine Science, vol. 64,no. 1, pp. 129–141, 1999.

[55] D. J. Skillings and R. J. Toonen, “It’s just a flesh wound: non-lethal sampling for conservation genetics studies. Proceedingsof the American Academy of Underwater Sciences,” in Pro-ceedings of the 29th American Academy of Underwater SciencesSymposium: Diving for Science, N. W. Pollock, Ed., AAUS,2010.

[56] T. A. Ebert, “Recruitment in echinoderms,” in EchinodermStudies 1, M. Jangoux and J. M. Lawrence, Eds., A.A. Balkema,Rotterdam, The Netherlands, 1983.

[57] J. Lee, M. Byrne, and S. Uthicke, “The influence of populationdensity on fission and growth of Holothuria atra in naturalmesocosms,” Journal of Experimental Marine Biology andEcology, vol. 365, no. 2, pp. 126–135, 2008.

[58] S. Rozen and H. Skaletsky, “Primer3 on the WWW for generalusers and for biologist programmers,” in Bioinformatics Meth-ods and Protocols: Methods in Molecular Biology, S. Krawetzand S. Misener, Eds., pp. 365–386, Humana Press, Totowa, NJ,USA, 2000.

[59] J. D. Thompson, D. G. Higgins, and T. J. Gibson, “CLUSTALW: improving the sensitivity of progressive multiple sequencealignment through sequence weighting, position-specific gappenalties and weight matrix choice,” Nucleic Acids Research,vol. 22, no. 22, pp. 4673–4680, 1994.

[60] T. A. Hall, “BioEdit: a user-friendly biological sequence align-ment editor and analysis program for Windows 95/98/NT,” inNucleic Acids Symposium Series, vol. 41, pp. 95–98, 1999.

[61] H. J. Bandelt, P. Forster, B. C. Sykes, and M. B. Richards,“Mitochondrial portraits of human populations using mediannetworks,” Genetics, vol. 141, no. 2, pp. 743–753, 1995.

[62] H. J. Bandelt, P. Forster, and A. Rohl, “Median-joiningnetworks for inferring intraspecific phylogenies,” MolecularBiology and Evolution, vol. 16, no. 1, pp. 37–48, 1999.

[63] M. Nei and W. H. Li, “Mathematical model for studyinggenetic variation in terms of restriction endonucleases,”Proceedings of the National Academy of Sciences of the UnitedStates of America, vol. 76, no. 10, pp. 5269–5273, 1979.

[64] J. Rozas, J. C. Sanchez-DelBarrio, X. Messeguer, and R. Rozas,“DnaSP, DNA polymorphism analyses by the coalescent andother methods,” Bioinformatics, vol. 19, no. 18, pp. 2496–2497,2003.

[65] L. Jost, “GST and its relatives do not measure differentiation,”Molecular Ecology, vol. 17, no. 18, pp. 4015–4026, 2008.

[66] F. Tajima, “Statistical method for testing the neutral mutationhypothesis by DNA polymorphism,” Genetics, vol. 123, no. 3,pp. 585–595, 1989.

[67] Y. X. Fu, “Statistical tests of neutrality of mutations againstpopulation growth, hitchhiking and background selection,”Genetics, vol. 147, no. 2, pp. 915–925, 1997.

[68] L. Excoffier, G. Laval, and S. Schneider, “Arlequin version 3.0:an integrated software package for population genetics data


analysis,” Evolutionary Bioinformatics Online, vol. 1, pp. 47–50,2005.

[69] Y. Benjamini, A. M. Krieger, and D. Yekutieli, “Adaptive linearstep-up procedures that control the false discovery rate,”Biometrika, vol. 93, no. 3, pp. 491–507, 2006.

[70] I. Dupanloup, S. Schneider, and L. Excoffier, “A simulatedannealing approach to define the genetic structure of popula-tions,” Molecular Ecology, vol. 11, no. 12, pp. 2571–2581, 2002.

[71] F. Manni, E. Guerard, and E. Heyer, “Geographic patternsof (genetic, morphologic, linguistic) variation: how barrierscan be detected by using Monmonier’s algorithm,” HumanBiology, vol. 76, no. 2, pp. 173–190, 2004.

[72] K. Tamura and M. Nei, “Estimation of the number ofnucleotide substitutions in the control region of mitochon-drial DNA in humans and chimpanzees,” Molecular Biologyand Evolution, vol. 10, no. 3, pp. 512–526, 1993.

[73] D. Posada and K. A. Crandall, “MODELTEST: testing themodel of DNA substitution,” Bioinformatics, vol. 14, no. 9, pp.817–818, 1998.

[74] F. Ronquist and J. P. Huelsenbeck, “MrBayes 3: Bayesianphylogenetic inference under mixed models,” Bioinformatics,vol. 19, no. 12, pp. 1572–1574, 2003.

[75] S. Tavare, “Some probabilistic and statistical problems in theanalysis of DNA sequences,” in Lectures on Mathematics inthe Life Sciences, vol. 17, pp. 57–86, American MathematicalSociety, Providence, RI, USA, 1986.

[76] A. Stamatakis, “RAxML-VI-HPC: maximum likelihood-basedphylogenetic analyses with thousands of taxa and mixedmodels,” Bioinformatics, vol. 22, no. 21, pp. 2688–2690, 2006.

[77] M. A. Miller, M. T. Holder, R. Vos et al., “The CIPRES Portals,”2010, http://www.phylo.org/sub sections/portal/.

[78] P. Beerli, “Comparison of Bayesian and maximum-likelihoodinference of population genetic parameters,” Bioinformatics,vol. 22, no. 3, pp. 341–345, 2006.

[79] M. Slatkin and N. Barton, “A comparison of three indirectmethods for estimating average levels of gene flow,” Evolution,vol. 43, pp. 1349–1368, 1989.

[80] D. W. Hutchison and A. R. Templeton, “Correlation of pair-wise genetic and geographic distance measures: inferring therelative influences of gene flow and drift on the distributionof genetic variability,” Evolution, vol. 53, no. 6, pp. 1898–1914,1999.

[81] M. Slatkin, “Gene flow and the geographic structure of naturalpopulations,” Science, vol. 236, no. 4803, pp. 787–792, 1987.

[82] S. E. Gilman, “Life at the edge: an experimental study of apoleward range boundary,” Oecologia, vol. 148, no. 2, pp. 270–279, 2006.

[83] M. T. Craig, J. A. Eble, B. W. Bowen, and D. R. Robertson,“High genetic connectivity across the Indian and PacificOceans in the reef fish Myripristis berndti (Holocentridae),”Marine Ecology Progress Series, vol. 334, pp. 245–254, 2007.

[84] J. K. Schultz, R. L. Pyle, E. DeMartini, and B. W. Bowen,“Genetic connectivity among color morphs and Pacificarchipelagos for the flame angelfish, Centropyge loriculus,”Marine Biology, vol. 151, no. 1, pp. 167–175, 2007.

[85] M. R. Gaither, R. J. Toonen, D. R. Robertson, S. Planes, andB. W. Bowen, “Genetic evaluation of marine biogeographi-cal barriers: perspectives from two widespread Indo-Pacificsnappers (Lutjanus kasmira and Lutjanus fulvus),” Journal ofBiogeography, vol. 37, no. 1, pp. 133–147, 2010.

[86] J. A. Eble, R. J. Toonen, and B. W. Bowen, “Endemism anddispersal: Comparative phylogeography of three surgeonfishesacross the Hawaiian Archipelago,” Marine Biology, vol. 156, no.4, pp. 689–698, 2009.

[87] A. Laxminarayana, “Induced spawning and larval rearing ofthe sea cucumbers, Bohadschia marmorata and Holothuria atrain Mauritius,” Fisheries Research, pp. 48–52, 2005.

[88] C. Riginos, K. E. Douglas, Y. Jin, D. F. Shanahan, and E. A.Treml, “Effects of geography and life history traits on geneticdifferentiation in benthic marine fishes,” Ecography. In press.

[89] J. E. Maragos and P. L. Jokiel, “Reef corals of Johnston Atoll:one of the world’s most isolated reefs,” Coral Reefs, vol. 4, no.3, pp. 141–150, 1986.

[90] C. E. Bird, B. S. Holland, B. W. Bowen, and R. J. Toonen,“Contrasting phylogeography in three endemic Hawaiianlimpets (Cellana spp.) with similar life histories,” MolecularEcology, vol. 16, no. 15, pp. 3173–3186, 2007.

Submit your manuscripts athttp://www.hindawi.com

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation http://www.hindawi.com

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttp://www.hindawi.com

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research


International Journal of

Microbiology

Documents

GatewaystoHawai‘i:GeneticPopulationStructureoftheTropical ...downloads.hindawi.com/journals/jmb/2011/783030.pdf · Kingman reef are important centers of the genetic diversity in